sign in sign up
<<
Home , Comparative statics, John Hicks, John Hicks (pianist), Value and Capital

A of

Leibniz-Informationszentrum econstor Wirtschaft Leibniz Information Centre Make Your Publications Visible. zbw for

Leijonhufvud, Axel

Working Paper Hicks on time and

Diskussionsbeiträge - Serie A, No. 182

Provided in Cooperation with: Department of Economics, University of Konstanz

Suggested Citation: Leijonhufvud, Axel (1984) : Hicks on time and money, Diskussionsbeiträge - Serie A, No. 182, Universität Konstanz, Fakultät für Wirtschaftswissenschaften und Statistik, Konstanz

This Version is available at: http://hdl.handle.net/10419/75126

Standard-Nutzungsbedingungen: Terms of use:

Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Documents in EconStor may be saved and copied for your Zwecken und zum Privatgebrauch gespeichert und kopiert werden. personal and scholarly purposes.

Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle You are not to copy documents for public or commercial Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich purposes, to exhibit the documents publicly, to make them machen, vertreiben oder anderweitig nutzen. publicly available on the internet, or to distribute or otherwise use the documents in public. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, If the documents have been made available under an Open gelten abweichend von diesen Nutzungsbedingungen die in der dort Content Licence (especially Creative Commons Licences), you genannten Lizenz gewährten Nutzungsrechte. may exercise further usage rights as specified in the indicated licence. www.econstor.eu Universitat Konstanz

A

Fakultat fiir Wirtschaftswissenschaften und Statistik

AxeljJ_eijonhufvud

Hicks On Time And Money

Diskussionsbeitrage

Postfach 5560 SerieA— Nr. 182 D-7750 Konstanz Mai 1984 •i JUL1 19B'i HICKS ON TIME AND MONEY

Axel jLeijonhufvud *

Serie A - Nr. 182

Mai 1984

•Member (1983-1984), Institute for Advanced Study, Princeton. Professor of Economics, University of California, Los Angeles Standiger Gastprofessor, Universitat Konstanz. •'*

Serie A: Volkswirtschaftliche Beitrage

Serie B: Finanzwissenschaftliche Arbeitspapiere

Serie C: Betriebswirtschaftliche Beitrage My assignment is Hicks and Keynes. It is too large for a paper: Modern macroeconomic theory has been shaped to an extraordinary degree by these two men. I .will con-fine my discussion of Hicks's role to two related themes: Time and Money.

Even within these boundaries, the fallowing attempt at an 1 interpretation cannot be definitive. Among the several reasons •for this, one is germane: I know that I will learn more -from Sir

John Hicks in the future. But I cannot know exactly what I will learn newt time I sit down to read or reread him. Hence today's assessment cannot be my "optimal" or final one. Rather than commit myself fully, I should retain a measure of "flexibility."

In certain types of situations, it is rational to commit oneself fully or contingently. In others, where the future contingencies cannot be enumerated or their nature anticipated, one should retain flexibility. One difference between neoclassical and Keynesian theory is that the former tends to exclude, whereas the latter must include, situations of the 3 second sort. The younger Hicks is remembered for his contributions to ; over the years the elder

Hicks has become more insistently Keynesian in this particular sense. "Every is familiar with the accomplishments of

Hicks the Younger, whether he has read him or not. That brilliant young man was supremely successful — by reformulating theory, by simplifying monetary theory, by interpreting Keynes and the Glassies, and by reviving general equilibrium theory — in constructing the molds into which 40 years of subsequent 4 theoretical developments were to be cast." It is helpful to try to see the young Hicks in historical context.

What went on at the London School in the early thirties appears in retrospect almost as important as what was going on in

Cambridge. At LSE, the world of Anglo-American economics was being won over from the tradition of Ricardo and Marshall to modern neoclassical economics—or, in the terms of Hicks the

Elder, from "plutology" to "catallactics." If Cambridge was sufficient unto its British self, Lionel Rob'bins's London School encouraged the study of the Austrian and the Lausanne schools, of the Americans and the Swedes. ("We were such 'good Europeans' in 5 London that it was Cambridge that seemed 'foreign'.") Robbins brought Hayek to London and assembled a stable of superbly talented junior people: R.6.D. Allen, Marian Bowley, and Ursula Webb-Hicks, , Abba Lerner, ^'era. Smith-

Lutz, Richard Sayers and G.L.S. Shackle. Most importantly,

Robbins wrote the programmatic tract that, highly controversial

in its time, has long since permeated the teaching of economics

to the point where its main message has become a platitude (thus

depriving its author o-f the ?) His Nature and of Economic Science argued the "" definition of economics, a definition that fundamentally changed both the scope and the content of Marshall's subject. Robbins made rational means-ends calculation the core of economics. v

It was the younger Hicks that demonstrated how this Robbins program could be realized. The Hicks-Allen "Reconsideration* recast demand theory in terms of rational .

Hicks's simplification of monetary theory drew Money into the orbit of marginalist calculation. "Taking step after step along a road which seemed pre-ordained as soon as one had taken the first

step" in a few years time led to the 'static' parts (Chapters I- 6 VIII). of Va^ue and .. These were the parts of Hicks's early work that, together with "Keynes and the Classics", were to

have such a profound and pervasive influence on how economics was

to be taught in the United States in the era when American

economics was becoming strongly predominant. Perhaps it is more

accurate to say that these parts of Hicks's work were selected by

the generation of American led by .Paul Samuel son that

were reerecting the structure of economic theory using

constrained optimization building blocks.

Pure decision theory, formalized as optimization subject to

constraints, is essentially timeless. The choice among the 7 foreseen outcome's of alternative actions is a purely logical calculus that does not involve time in any essential way. Thus

was created a durable tension between neo-Walrasian microtheory

and Keynesian macrotheory that, decades later, was to culminate

in crisis.

This could hardly have been foreseen. As Robert Clower has 8 remarked,

... it was only natural for economists generally to proceed on the presumption that general equilibrium theory had no inherent limitations.... That any even moderately "general" should Cbe incapable of representing Keynesian processes]...would hardly occur naturally to any but a very perverse mind. That the elaborate Neo-Walrasian model set out in Hicks' Val_ye and Capital, might fail tin this respect] would have seemed correspondingly incredible to any sensible person at the outset of the Neo-Walrasian Revolution.

The younger Hicks knew that Time was a problem. We find him

wrestling with it in almost all the parts of his early work that

did not become part of the American neoclassical canon. It was to

become even more of a preoccupation — an unfashionable

preoccupation — for Hicks the Elder.

From the first, it seems, Hicks saw it as a supreme

theoretical challenge, deserving the most sustained effort, to

find a mode of process analysis that would retain a role for

equilibrium constructions without denying ', (or trivializing)

change. In the early going, this amounted to finding a workable

way between Walras and Pareto, on the one hand, and Knight and 9 Hayek on the other. Thirty or forty years later, the opposed alternatives — Arrow-Debreu vs. Shackle or Lachmann —_^

clearer and also further apart. Shackle poses the issue with

uncompromising force: "... the theoretician is confronted with a 10 stark choice. He can reject rationality or time." The American Neo-Walrasians, from Paul Samuel son to Robert

Lucas, have not seen this choice as at all difficult. In general,

they have simply gone whole hog for Rationality, letting Time and

Change be trampled underfoot in the philosophical muck as unfit food for economic thought. If forced (somehow) to choose, -it is possible that Hicks the Younger might also have opted for rational allocation theory; Hicks the Elder almost certainly would opt for . In actuality, Hicks fought fifty years to maintain a conceptual middle ground.

The issue may have come into focus at LSE precisely because all of the neoclassical schools were to some extent cultivated in the circle around Robbins and Hayek. Marshall had been aware of 11 the problem and had devised a method that at least partly evaded it. Hayek had worked on the construction of an equilibrium process "in time" and had found himself forced back onto 'perfect 12 foresight' assumptions. Robbins had drawn the conclusion that

"(t)he main postulate of the theory of dynamics is the fact that 13 we are not certain regarding future ."

As matters stood around 1930, the static toolbox of economic theory was strictly applicable only to stationary, perfect foresight processes. It was not at all clear that economic theory provided any foundation for the disciplined analysis of monetary questions or business cycles. Hicks's earliest work dramatized the predicament. In particular, his remarkable 1933 paper on 14 "Equilibrium and the Cycle" drove home a point made by Knight: that in a perfect foresight equilibrium process, people would not demand cash-balances. This spelled trouble for the most sophisticated cycle theory available at the time. What became of

Hayek's notion of "neutral money" as a criterion for maintaining macroeconomic equilibrium, if in equilibrium there could be no place for money, "neutral" or otherwise?

The Swedish followers of Wicksell had run into similar quandaries and it was from Myrdal and Lindahl that Hicks got help 15 with the next step. The next step had to ""be a method of describing economic processes that (a) was not confined to just

'perfect foresight' processes, and (b) still did not force the abandonment of the entire apparatus of inherited static theory,, 16 Lindahl's temporary equilibrium method "reduced the process of change to a sequence of single periods, such that, in the interior of each, change could be neglected.... Everything is just the same as with the 'static' kind of process analysis ... save for one thing: that expectations are explicitly introduced as independent variables in the determination of the single-period equi1ibrium.

Thus, when the General, Ihegry. appeared, Hicks had been working

along these lines for some time. His first reaction gave pride of

place to Keynes's use of a similar device: a short-run

equilibrium adapting to independently specified lonq-term 17 expectations. But the kinship was not all that close. Keynes had applied the "method of expectations" to a Marshallian short

period. Marshall had invented a kind of analysis ("with some 18 slight dynamic flavoring" ) which definitely was "in time" but

that left the line between statics and dynamics unclear. In

and Capital., Hicks developed an alternative line of attack.

The attack starts with the famous definition of "Economic

Dynamics" as those parts of economic theory "where every quantity 19 must be dated." This was an important step. The Marshal 1ians,

for example, had not taken it.

By itself, the dating of only adds dimensions to the

commodity space considered in "timeless" statics. Studies in

efficient intertemporal resource allocation following Fisher and Hicks have improved our understanding of capital, growth and theory immensely. But the course of this development became quite similar to what happened to British Classical theory about which Hicks observed: "The more precise capital * theory became, the more static it became; the study of equilibrium 20 conditions only resulted in the study of stationary states." We have to substitute "steady" for "stationary", of course, but

otherwise the conclusion holds. It is presumably for this reason

that Hicks no longer favors his old static-dynamic distinction

but prefers to talk of analysis that is "out of time" or "in 21 time." Dating brings in future time, but it does not necessarily

help in bringing in the passage of time. If the usual

(stochastically) perfect knowledge assumptions are made, the end

result will be the Arrow-Debreu contingency model in which

all decisions are made at the origin of time. There is no-

business left to transact at later dates. Money and liquidity can

be forced into such a structure only by obvious artifice.

The present-day practice at this juncture is for the

theorist to retire behind a smoke screen while intoning some

incantation about transactions costs. Hicks, in 1939, did a~bit

better. What must be done is to weaken the informational

assumptions of the model so as to make agents postpone at least 22 some decisions "until they know better." Hicks discussed several types of uncertainty and decided, I think correctly, that

agents' uncertainty about their own intentions was the most 23 fundamental:

... in particular, they know that they cannot foretell at all exactly what quantities they will themselves desire to buy or sell at a future period.... and this it is, in the end, which limits the extent to which forward trading can be carried on in practice. .;k This argument is the bridge by which Hicks made his escape from steady-state capital theory into temporary equilibrium theory. In the temporary equilibrium theory of Val_ue and Capital., time is divided into a sequence of "weeks." Planned demands and supplies for the week depend on current and expected future prices. Current prices are determined on "Monday" and rule unchanged for the rest of the week. On "Sunday" (we may imagine), the parameters of the equilibrium system are updated: changes in stocks are accounted for and —expectations revised. The system is then ready for another Monday morning.

In this story, all markets cleared each Monday. Hicks understood perfectly that this assumption by itself did not preclude periods of subnormal activity in the system. The defense of the assumption that he suggested is exactly the one so strenuously insisted upon by Lucas, Barro et al.ia almost forty years later. In Hicksian terms, if price-expectations are inelastic, a fall in current prices will induce intertemporal substitution: supplres will be shifted from this week i.nto 24 next. Market clearing, however, was equilibrium in a "limited sense"; in the more fundamental sense of "Equilibrium over Time",

Hicks emphasized, the was "usually out of 25 equi1ibri urn."

This Temporary Equilibrium method is thus clearly distinct both from Keynes' short-run equilibrium, on the one hand, and from the New Classical equilibrium method of more recent years,

8 on the other. It avoids some of the problems of the alternatives 26 and deserves further exploration, therefore, although of course, it has problems of its own. But, while Hicks has resumed the struggle for a systematic "in time" analysis later — and'on more 27 difficult ground even — he chose to abandon the Temporary

Equilibrium approach.

Why? The Elder Hicks has given his retrospective reasons. 28 There were problems within the 'week' and between 'weeks':

Much too much had to happen on that 'Monday'! And...I was really at a loss how to deal with the further problem of how to string my 'weeks* and my 'Mondays' together.

Getting from one 'week' to the next required both a theory of and a theory of the revision of expectations. The first problem by itself was forbidding at the time; only the later development of modern growth theory made it manageable. Forty years have not brought us much advance on the 29 second problem. In his retrospective evaluation, Hicks does not point to the> problems that the Temporary Equi1ibrium'method would have to overcome in order to provide a 'continuation' theory; instead, he ' 30 focuses on how the method dealt with events 'within the week':

... I • tried to go further Cthan to work with ^ expectations'], to allow for the effects of current transactions on expectations; supposing that these effects could (somehow) be contemporaneous with the transactions themselves.... That however was nonsense....'It does deliberate violence to the order in which in the real world (in any real world) events occur.' It was this device, this indefensible trick, which ruined the 'dynamic' theory of .. It was this that led it back in a static, and so in a neo-classical, directi on.

What an extraordinarily harsh judgment this is! Why? Because in obliterating the seguence in which things happen, the model comes to ignore the structure of markets. It matters, for instance, whether people commit themselves*on quantities and discover their mistakes through price—change 'surprises* or set their prices and v 31 see their errors revealed in the behavior of quantities. It matters, in Hicks' terms, whether the markets in the system are

mostly of the fLexzECi^S. or the fi^zECics variety. In this

century, "the unorganized flexprice market, the old type, is on

the way out .... modern markets are predominantly of the fixprice 31 type..." In Hicks' view, this historical transformation is of major macroeconomic significance. The change in the predominant

market form is a change in the way that impulses are propagated

through the system. The harsh language becomes understandable —

for, of course, Hicks sees the "indefensible trick" still being

practiced all over!

The younger Hicks may have had somewhat different reasons

for abandoning his Temporary Equilibrium method. One of them

surely was that Keynes had come up with an alternative method of

short-period analysis., It was a rough-and-ready sort of short-

period method and Hicks the Younger would have realized better

than anybody else how rough it was. But it seemed to be adequate

to Keynes' purposes and Hicks agreed that Keynes* purposes were

the supremely important ones.

Soon after his original review. Hicks returned to the

@§D§C£l Ib-Sory. and wrote "Mr. Keynes and the 'Classics': A

Suggested Interpretation." The IS-LM apparatus of this immensely

10 influential paper was not a Walrasian (or Paretian) construction but a hybrid. Keynes' macrotheory was built-- with Marshal li an microcomponents. But the modelling idea was, as Hicks has himself 33 . * explained, borrowed from Value and Capital, where he hads worked out a two-dimensional representation of the equilibrium for a

Walrasian'system of three markets.

The IS-LM model summarized numerous features of the General

Ib-gory. with admirable and it was to serve in the

deduction of numerous Keynesian propositions

that Keynes had not thought of. The model became the backbone of

instruction in for forty years. Nonetheless,

something was just a bit askew with it. In later years, Hicks has

several times come back to reassess it and the uses to which it

has been put. In brief, (a) he has remained fairly content with

it as a synopsis of Keynes* theory; (b) he has become less

satisfied with it as a way of portraying the*'Classics* and hence

as a tool for isolating Keynes' contribution by IS-LM 35 * comparisons; and (c) he has grown somewhat skeptical about it 36 as a general purpose framework for macroeconomic analysis. His several commentaries on IS-LM all focus on the problem of Time. From the early fifties to the mid-sixties, Hicks did not 37 participate much in ongoing developments in economic theory.

When he returned to theoretical work full time, he was eager to

learn what had been accomplished in growth theory but found

himself out of sympathy with the directions taken in

macroeconomics and monetary theory. The trouble was that these

directions had been set by Hicks the Younger — in those parts of

11 his work that the American economists had chosen to cultivate.

Hicks' first dismaying confrontation with his own brain-children

— now fully grown and so independent! — came, it appears, in

1957 when he was asked to review Patinkin's first edition.

Patinkin's work had been systematically and rigorously built on the basis of the Hicks-Allen 'Reconsideration', the paper

'Simplifying* monetary theory, 'Keynes and the Classics*, and the first eiqht chapters of Value and Capital (toqether with some "38 closely related works by Oscar Lange). But the theoretical structure that Patinkin had erected on these foundations, Hicks 39' thought, threatened to emasculate . Never a whole-hearted Keynesian, Hicks was nonetheless too much of a

Keynesian to stand, idly by under the circumstances.

Patinkin's basic model was a Walrasian general equilibrium' model, built up from choice-theoretical Individual experiments," via aggregation, to equilibrium market experiments. It allowed no

Marshallian distinctions between short-run and long-run equilibria. It was either in "the" equilibrium or not in equilibrium at all. Patinkin used the Hicksian technique for portraying the equilibrium of an aggregative version of the system as the intersection of two reduced forms in

interest/ space. It "looked" exactly like IS-LM — except 40 this version would not allow for .

Hi-cks set out to show that 'Classical' and Keynesian theory

"do not overlap all the way" — that all the Keynesian furor had 41 not been pointless. His point of departure was the right one: The crucial point, as I now feel quite clear, on which the individuality of the Keynes theory depends, is the implication ... that there are conditions in which the

12 -4 JUNt 1385 ««£f* interest—mechanism will not work.

In the original Patinkin review, Hicks tried to show this in two ways. His first argument, however, amounted to a reassertion of v the explanation of unemployment and Patinkin had only to repeat his demonstration of how, with flexible , the

Pigou-effect would restore full . Within the IS-LM context, the explanation of unemployment is thus thrown back unto 42 the "rigid wages" postulate. Hicks' second and surviving argument attempted to clarify the relationship between Keynes and the 'Classics' by showing how the parameters of the IS-LM model depend on the length of period assumed. The extent to which wages are variable, Hicks pointed out, will depend not only on the magnitude of excess demand (or supply) of labor but also on the

length of time allowed for adjustment. Over a sufficiently , long

period,' the IS-schedule should then be infinitely elastic (at the

'natural rate' of interest), while the speculative component

disappears from money demand so that the LM-schedule becomes

quite inelastic. With a shorter period, the 'Classical' dichotomy

fails, and the shorter the period the more 'Keynesian' the

picture: IS becomes very inelastic and LM exceedingly elastic in 43 the very short run.

This defense of Keynes (if such it was) could only focus

attention on Keynes' own treatment of time, however. Hicks'

reservations on this score (as well as those of other "critical

readers") went back all the way to the thirties: "... but we have

agreed to suspend our doubts because of the power of the analysis 44 which Keynes constructed on this (perhaps) shaky foundation." It could not be left at that indefinitely. In his 1974 effort to

13 address The Crisis in Keynesian Economics, Hicks left the matter to one side and simply made no us^e of IS-LM at--~all. But in

Ib.§ SclliSj he advanced the theory of liquidity as flexibility as one of the needed cures for the ailing Keynesian tradition. In 45 contrast to how it emerges in static portfolio theory,

... 'liquidity is not a property of a single choice? it is a matter of a sequence of choices, a related sequence. It is concerned with the passage from the known to the unknown •— with the knowledge that if we wait we can have more knowledge. So it is not sufficient, in liquidity theory, to make a single dichotomy between the known and the unknown. There is a further category, of things which are unknown now, but will become known in time.

This, clearly, lends urgency to the question of how time is to be treated in Keynesian models. Immediately afterward, therefore.

Hicks turned to reexamine the compromises of Keynes' method and 46 found them, on close inspection, less and less satisfactory: Keynes's theory has one leg which is in time, but another which is not. It is a hybrid. I am not blaming him for this; he was looking for a theory which would be effective, and he found it.... but what a muddle he made for his successors!

In brief, the "leg in time" is LM, the "leg in equilibrium" is

IS. (Clearly, this "straddle", as Hicks called it, was a position that had to become uncomfortable with the passage of time!) 47 Hicks' own Temporary Equilibrium method also was divided; there was a part that was in time and a part that was not. But we did not divide in the same place. While Keynes had relegated the whole theory of production and prices to equilibrium economics, I tried to keep production in time, just leaving prices to be determined in an equilibrium manner.

Production will not be equilibrated in a 'week'. Hicks' 1983 "IS- 48 LM: An Explanation" carries the argument forward:

If one is to make sense of the IS-LM model while paying proper attention to time, one must, I think, insist on two things: (1) that the period in question is a relatively long

14 period, a 'year' rather than a 'week'; and (2) that ,because the behaviour of the economy over that 'year' is to be determined by propensities and such-like data, it must be assumed to be, in an appropriate sense, iD

Product markets are in f_lgw equilibrium throughout the \year'; production plans are being carried through without disappointment or surprise; this, in Hicks' view, is how we must interpret the

IS-curve. What about the LM-curve? It is.a stock-relation and, by itself, could apply simply to a point in time. But to be consistent with the IS-construction, Hicks paints out, a more restrictive equilibrium condition should be applied, namely, maintenance of stock equilibrium throughout the 'year'.

Expectations and realizations must be consistent within the period. But at this point of his 1983 argument, we are suddenly back facing the dilemma of that 1933 paper: "Disequilibrium is the Disappointment of Expectations" — and in equilibrium processes there is no place for money! The "Equilibrium method, 49 applied to liquidity over a period, will not do." Within the IS-LM construction itself, therefore, we find this tension between Equilibrium and Change which I see as a

Leitmotif through five decades of Hicks' work. Hicks is "quite prepared to believe that there are cases" where we are "entitled to overlook" the potential inconsistency between the ways that the IS and the LM have been constructed. But he clearly no longer regards it as a robust tool for the analysis of almost all 50 macroeconomic questions.

IS-LM served us well for so long (didn't it?). How could we not have run into obvious problems with it, if it teeters on the brink of conceptual inconsistency? IS-LM exercises produce the right answers (most of us will agree) to -a... large number of standard macroquestions. Yet, it produces the wrong conclusions y (some of us insist) on some issues. Hicks' leaves us with a general skepticism about the method which does not help us much

in determining what uses are safe and what uses are not.

In an attempt to find out "What was the Matter with IS-LM?",

I came to a conclusion very similar to Hicks' judgment on the

Temporal Equilibrium method: As with all equilibrium

constructions, IS-LM ignores the segyence of events within the 51,52 period. The result can be nonsense: IS-LM, handled as if it were a static construction .... produced a nonsensical conclusion to the Keynes and the classics debate: namely, that Keynes had revolutionized economic theory by advancing the platitude that wages too high for full employment and rigid downwards imply peV-sistent unemployment. It failed to capture essential elements of Keynes's theory: namely, that the typical shock is a shift in investors' expectations and that it is the failure of intertemporal prices to respond appropriately to this change in perceived intertemporal opportunities that- prevents rational adaptation to the shock. The same 'as if static' method produced the conclusion that iisyidity Edeference versus loanable funds was not a meaningful issue? that it does not matter whether the system is or is not potentially capable of adjusting intertemporal prices appropriately in response to changes in intertemporal opportunities. Ignoring sequencing becomes a source of trouble in

particular in connection with comparative statics uses of the IS-

LM model — i.e.,the uses that are the stuff which macrotexts

have been made of for several decades, but which Hicks did not

consider in reassessing the model.

Consider, for illustrative purposes, the analysis of an

increase in the supply of money in the common textbook context

16 where the is simply an exogenously fixed M. Full adjustment to this parametric disturbance requires a proportional rise in all money prices with no effect on , employment or other real magnitudes. In an IS—LM diagram with money income on the horizontal axis, both, schedules have shifted the same distance rightwards. In a Lucas model, if M is observable, the system goes to this position immediately. In a Friedman model

(of, say, 10 years ago), on the other hand, nominal income responds strongly in relatively short order, but part of this is an increase in real output and employment and full adjustment to the neutral equilibrium takes "longer." In a Keynesian model (of

20 years ago?), finally, the "short run" reactions are that the falls, velocity declines and and employment increase a bit.

All' three possibilities can be demonstrated with the same basic model. How, then, do they differ? To Friedman, the.

Phillips-curve is vertical only over the "long run", not already in the "short run" as in Lucas. In Friedman's short run," the monetary disturbance has output effects because people temporarily miscalculate real wages. To the Keynesians, the

(approximately) proportional increase in nominal income occurs only over the "long run", not already in the "short run" as in

Friedman. In the Keynesian short run, the monetary disturbance has only weak effects on nominal income now because people fail to anticipate the effect that it must have on nominal sooner or later; hence the short run effects on income occur only in so far as some firms are induced by a fall in the

interest rate to increase their investment even though their

17 expectations of future nominal aggregate demand have not 53 improved. ,-4 So, Lucas' people are assumed to know something that

Friedman people do not, and Friedman people something that 54 Keynesian people do not. The temporal order of decisions matters when information is incomplete, when people have to react to situations they did not foresee and when they learn from realizations they did not anticipate. Such learning can be slow or fast or, in some cases, unnecessary.

Note how these knowledge or learning assumptions are reflected in the mechanics of manipulating the IS-LM diagram. In the Keynesian exercise, LM shifts right, IS stays put, and the short-run effects depend on the elasticities of the two reduced forms. In the Friedman case, LM also shifts, although perhaps not

Sylte all the way; the elasticities then are practically irrelevant. In the Lucas case, both reduced forms shift in parallell fashion. The IS-LM modelling strategy would seem to presuppose that we have to deal with a Keynesian world of slow learners. Otherwise it does not seem to make sense to adopt the two—stage procedure of, first, deriving the two reduced forms and, second, getting the answers by shifting one and keeping the other constant. The use of IS—LM as if it were a comparative static apparatus involves the lag-assumption that one schedule shifts before the other and that there will be a well—defined

"short run" solution halfway in the•equi1ibrating process. This sequencinq or lag structure rests on assumptions of incomplete 55 information on the part of various agents in the model.

18 This conclusion we have derived from an illustrative case where monetarist assumptions are made about the supply of money.

There is, however, also another possible interpretation of

Keynesian IS-LM analysis which we will come to later.

Q2Q§y £D. History In the most exciting chapter of his Critical Essays in d°D§tary Ib_egrv, Hicks sought to structure two centuries of monetary writings in a simple, striking, and informative way. His

"Monetary Theory and History — An Attempt at Perspective" was critical of ahistorical monetary theorizing and insisted on the necessity of doing monetary theory in historical and institutional context. It also suggested that the history of monetary controversy could be understood as a running battle between two traditions, a "metallic money" tradition and a

"credit money" tradition.

The 'metallic money' theorists, in Hicks' schema, focused on equilibrium propositions in their theorizing, dealt analytically with money "as if" it was a commodity, and strove to reduce to obedience to some "mechanical rule." Credit theorists, on the other hand, saw money as part of the overall system of debits and credits that extends beyond the banking system to encompass the entire economy; credit expansions and contractions were central to their conception of the subject and so obliged them to try their luck at disequilibrium analyses; always aware that credit rests on confidence, finally, writers in this tradition saw monetary policy as an exercise in judgment of contemporary conditions. Hicks named Ricardo the patron saint of

19 the 'metallic' tradition and gave Thornton the same status in the

'credit' school of thought. He saw the School and, later, Hayek, Pigou, Rueff and Friedman as Ricardo's followers and put the Banking School, Bagehot, Wicksell, Hawtrey,

Robertson, and Keynes in line of descent from Thornton.

In insisting on the close link between monetary theory and history. Hicks thought above all of the evolution of credit markets and financial institutions: "In a world of banks and insurance companies, money markets and stock exchanges, money is quite a different thing from what it was before these 56 institutions came into being." The metallic money theorists (including the modern monetarists) seemed determined to ignore this historical development. Consequently, Hicks' analysis suggested, time had put an ever-increasing distance between their \ 57 theory and reality. The 1967 "Perspective" helps one understand what Hicks regards as the important themes runninq through his own 58 contributions to monetary the.ory. Consider, once again, what aspects of the work of Hicks the Younger came to be influential and what aspects ignored. For decades, all graduate students have learned that the modern choice-theoretical money demand function stems from his 1935 'Simplifying' paper. Most will know that

Hicks already had the depending on wealth, on anticipated yields on alternative placements, and on the cost of asset transactions. Some may recall that his analysis was anything but reassuring on the stability of the function in terms of these arguments. Few (I am guessing) will remember that, in

20 Hicks' hands, the theory immediately suggested the beginnings of a theory of financial structure, of the composition of balance sheets and of intermediation. ^Balance sheet equilibria, he noted, v

CareD determined by subjective factors like anticipations, instead of objective factors like prices, CwhichD means that this purely theoretical study of money can never hope to reach -results so tangible and precise as those which in its more limited field can hope to attain. If I am right, the whole problem of applying monetary theory is largely one of deducing changes in anticipations from the changes in the objective data which call them forth. Obviously, this is not an easy task, and, above all, it is not one which can be performed in mechanical fashion.

In our textbooks, Hicks' paper is remembered for a money demand function with which any . latter-day monetarist could be comfortable. But, clearly, he was in the Credit tradition from the beginning!

Moreover, it is the neglected themes of Hicks the Younger that the Elder has taken up and carried forward. The first step beyond his 1935 position, came three decades later with the sketch in Capital and Grgwth C1965] of a simple financial system, 60 consisting of a bank, savers, and firms: Savers can hold their assets in bank money, or in securities (loans or equities) of the producing firms;... Firms have real assets, and they may have bank money; they have debts to ^the bank , and to the savers. The bank has debts owing- to it from the firms; it owes debts (bank money) to the firms and to the savers.

The "Two Triads" of 1967 introduced the classification of assets into running assets, reserve assets, and investment assets; the specific assets that served these functions would differ between the balance sheets of , of firms, and of banks; for each type of transactor, the three classes of assets could be matched up with Keynes' Transactions, Precautionary, and

21 Speculative motives; in Hicks' treatment, however, these three were no longer just motives for holding money but for preferring balance sheets of a certain structure. In "Monetary Experience and the Theory of Money" C1977], the financial structure of

Keynes' world was envisaged as three concentric sectors: (1) a banking 'core' with monetary liabilities and financial securities as assets; (2) a financial 'mantle' owing financial securities and holding industrial securities; and (3) an outer 'industry' owing the industrial securities and holding the (hard crust of?) the economy's productive assets (and some financial assets and money). In the 1982 "Foundations of Monetary Theory," Hicks added to this * monocentr ic1* credit economy model, some analysis also of a 'polycentric* world of multiple central banks (and flexible exchange rates).

What do we get out of this 'Credit' approach that a monetarist for 'money' app-aratus would not provide with less trouble? Hicks, of course, uses his financial structure model routinely in the analysis of a broad range of questions. In my view, however, the significant advantage of his approach is that it gives a better picture of the financial and monetary consequences of 'real causes': a rise in the anticipated yields on real capital will change the configuration of balance sheets desired by the business, household, and banking sectors; the financing of investment will in part be intermediated by the banks; consequently, an increase in income due to a rise in marginal efficiency of capital will normally be associated not only with a rise in velocity but also with an endogenous increase in the money supply.

Hicks* insistence on linking monetary theory to monetary history has been echoed in recent"*years by' theorists who insist that we must link short-run monetary ^theory to monetary regimes. These modern writers, however, have come to their preoccupation with the conditional nature of monetary theory from an entirely different angle. Their concern has been to keep track, not of slowly evolving financial institutions and markets, but of rapidly changing nominal () expectations. A 'monetary regime' may be defined as a system of expectations that governs the behavior of the public and is sustained by the consistent behavior of the monetary 61 authorities. Since the short-run effects of particular policy- actions, for example, depend upon the expectations of the public, it follows that we need a different short-run macromodel for each monetary regime. A regime change occurs when the behavior rules followed by the monetary authorities change. This 'regime approach' directs our attention to the history of monetary standards, viewed as methods for controlling the level of nominal prices, and to the system of nominal expectations that would

(rationally) go with each such method. —

Historically, we find two basic but contrasting conceptions of how price level control can be accomplished. I have labelled them the 'quantity principle' and the 'convertibility principle', respectively. Briefly (and perhaps a bit too simply) we may say that the quantity principle dictates that the government should control the 'quantity of money" while the private sector sets the price level; the convertibility principle, in contrast, dictates that the government set the nominal price of some 'standard commodity* while the private sector determines^the quantity of 62 money. The logically tidiest version of the first would-be a fiat standard with flexible exchange rates, and of the second a commodity standard with 'hard money* still in circulation. Price expectations on the fiat standard are almost entirely a matter of beliefs about what the government might choose to do; price expectations on the commodity standard (conditional on the belief that the standard will be adhered to!) are almost entirely a matter of forecasting 'real' business developments.

The two contrasting systems give the extremes on a more or less continuous spectrum of monetary regimes. The last fifty-odd years have taken us from a position rather close to the commodity standard ^end (in 1929) all the way to the extreme fiat standard end. (after 1971). We could proceed to classify macrotheories according to the segment of the regime-spectrum over which they might claim validity.

This classification of theories according to control—regime differs from the Hicksian schema of metallic money theories vs. credit theories and may be a useful complement to it. This may be seen, for instance, by considering how the American monetarists fit into Hicks' schema. In a metallic money world, money is a produced commodity and thus not neutral; the price level is determined (in the long run) by the cost of producing the metal; the money stock is endogenous and not subject to policy control; the 'mechanical' policy rule is to maintain the metallic standard. The 'mechanical' rule of the monetarists is to fix the

24 growth rate of some 'M'; it is predicated on the beliefs that 'M' is neutral and controllable (and 'more or less' independent of endogenous real factors); the objec't is to control nominal income in the short run and the price level over the longer run; t fixed exchange rates are readily sacrificed to this end. When Hicks includes both Ricardo and Friedman in the same 'metallic* tradition these points of contrast are obscured (even as the contrasts between Ricardo and Friedman, on the one hand, and

Thornton and the Radcliffe Report, on the other, are brought into focus). Similarly, Hicks' has come to prefer Wicksell's "pure credit" model (of an economy without 'hard money') as his vehicle for . explaining the central theoretical message of the 'credit 63 tradition*. But to a monetarist audience, for instance, the main lesson of Wicksell's cumulative process is simply that, on a fiat standard, interest targeting of monetary policy produces

D.omina! instability. A model of a system where convertibility anchors the price level — and, therefore, anchors rational price expectations as well — does a better job of fitting credit as a real magnitude into monetary theory. It is easier, in such a model, to show both how banking policy can influence investment and employment via the price and volume of 'real' credit and how real income movements can influence the supply of nominal money via the demand for 'real' credit.

Keynesian theory, to take a case in point, seems suited to regimes that behave as if monetary policy were constrained by the requirements of external if not also internal convertibility. The real quantity of money varies endogenously over the cycle in such regimes, nominal price level expectations' should be inelastic, best reasons for studying the elder Hicks, in fact, is precisely that he is less a prisoner of the younger Hicks's constructions 66 than are most of us." Among the* lessons' that Hicks the Elder 67 would impress on us, I have tried to bring out two: v One must assume that the people in one's models do' not know what is going to happen, and know that they do not know just what is going to happen. As in history!

Monetary theory, especially, has to be developed "in time Cwith] future becoming present, and present becoming past, as time goes 68 on. And "it belongs to monetary history in a way that economic 69 theory does not always belong to economic history."

27 Notes

1. I have made one previous attempt. My "Monetary Theory in Hicksian Perspective" was written in 1968 but not published until 1981, at which time I was still reasonably content witHi the paper. Once it was in print my understanding of some of . the issues began to change — as I will explain below.

2. Cf. Hicks, The Crisis in Keynesian Economics, Chapter 2, and the antecedent Hart C19423.

3. In stressing this particular distinction between neoclassical and Keynesian theory over others, I am following G.L.S.Shackle more than my own earlier work. Cf. esp. Shackle C19723.

4. Quoting my own C19793 review of Hicks's Ecgngmic Perspectives C19773.

5. Cf., the "Commentary" to The Jhegry of Wages, C19633, p. 306. 'Plutology' and *catal1actics' are discussed in Hicks, "'Revolutions' in Economics," in Spiro Latsis, ed., C19763 reprinted in Hicks, Classics and Moderns, II19833.

£• ^£20.0.03.10 Eerspectiyes, pp. v-vi . 7. The foreseen consequences may of course be probability distributions of outcomes. This does not alter the problem.

8. Cf., Clower C19753, p. 134.

9. Cf. Hicks, "The Formation of an Economistr" Z1979b, p.1993, now reprinted in his C19833.

10. Cf. Shackle C19723, Preface.

11. Hicks, Capital and Grgwth, pp. 47-48 quotes Marshall C19283, p. 379, n. l: "A theoretically perfect long period ... will be found to involve the supposition of a stationary state of industry, in which the requirements of a future age can be anticipated an indefinite time beforehand.... and it is to this cause more than to any other that we must attribute that simplicity and sharpness of outline, from which the economic doctrines in fashion in the first half of this century derived some of their seductive charm, as well as most of whatever tendency they may have had to lead to false practical conclusions." :

Of course, the second hal'f of the 20th century takes a generally more permissive attitude to "seductive charms" than this most eminent Victorian among economists. Shackle's aptly titled chapter "Marshall's Accomodation of Time", in his C19723, gives a sample of other remarks of Marshall's indicating his preoccupation with the issue.

12. Cf. Hayek C19283. v 13. Robbins C19323, p. 79. 14. "Gleichgewicht und Konjunktur," Zeitschrift fur N.§tigna!gkgngmie, No. 4, 1933. This remarkably modern, historically important paper was finally translated and published in iogngmic lQ9.yiry, November 1980,thanks to its then editor, Robert Clower. It is now reprinted in Hicks C19823.

15. G. Myrdal, "Geldtheoretisches Gleichgewicht," in F.A. Hayek, ed., Beitrage zur Geidthegrie, 1933, was reviewed by Hicks in ^ogngmica, Nov. 1934. The review is reprinted in Hicks C19823. G.L.S. Shackle, also a member of the Robbins circle, testifies to the great influence and importance of Myrdal*s contribution in his C19673, Chapters 9 and 10.

Of Lindahl's temporary equilibrium concept. Hicks first learned through personal acquaintance. He has discussed temporary equilibrium methods repeatedly, e.g., in Value and Capital, esp. Chapters IX-X and XX-XXII, in "Methods of "Dynamic Analysis" C19563 now reprinted in Hicks C19823, and in Capital and Growth, Chapter VI.

16. Hicks C19653, p. 60.

17. Hicks' 1936 Ecgngmic Journal review is "reprinted in Mgneyj. Interest and Wages as "The General Theory: A First Impression."

18. Surely, Hicks was thinking of Marshall when (Value and Capital, p. 115-6) he declined to follow "the usual course of economists in the past ... and giveO one's static theory some slight dynamic flavouring, (so that) it can be made to lock much more directly applicable to the real world.... But it will still be quite incompetent to deal properly with capital and interest, or fluctuations, or even money..."

le and Capital, p. 115. 20. Capital and Grgwth, p. 47.

21. Cf. esp. his "Time in Economics", as reprinted in C19S23, e.g., p. 291: "(Steady State economics)... has encouraged economists to waste their time upon constructions that are often of great intellectual complexity but which are so much out of time, and out of history, as to be practically futile and indeed misleading. It has many bad marks to be set against it."

22. It is for this reason that I have proposed changing the Hicksian definition of dynamics to "those parts of economic theory where decisigns must be dated." Cf. Leijonhufvud C1983b3.

23. Value and Capital, p. 137. Of course, this way out -of the predicament ultimately requires us to formulate a theory of the behavior of agents who know that they are likely to "foresee their own wants incorrectly" (p. 134). This problem Hicks, did not tackle in 1939. It is in his Crisis in Keynesian Economics, Chapter II, thirtyfive years later, that we find it addressed.

Decisi on—making by agents who know that they will know better later (but don't know, even probabilistically, what it is they will learn) will not fit naturally into the usual constrained optimization apparatus. For a comprehensive attack on the problem, cf. Ron Heiner C19S33.

24. Cf. Value and Capital, p. 131: "There is a sense in which current supplies and current demands are always equated in competitive conditions. Stocks may indeed be left in the shops unsold; but they are unsold because people prefer to take the chance of being able to sell them at a future date rather than cut prices in order to sell them now. The tendency for the .current price to fall leads to a shift in supply from present to future. An excess of supply over demand which means more than this is only possible if the price falls to zero, or if the commodity is monopolized, or if the price is conventionally fixed. "

25. Value and Capital, loc.cit.

26. It took more than 30 years for the profession to catch on to what Hicks had been up to in 1939. Grandmont's survey (1975) shows how the crisis of Keynesianism, which was in part a crisis of Keynes' method, had produced a more profound appreciation of the difficulties that the Temporary Equilibrium approach had been designed to address. 4 27. The "Traverse" problem which Hicks set himself in Chapter XVI °f Q§Eitai and Grgwth and analyzed at length in Capital and lime adds a forbidding burden of capital theory to the difficulties discussed in the text.

28. Cf., "Time in Economics," in Hicks (1982), p. 290. In 1956, ("Methods of Dynamic Analysis"), Hicks distinguished between the problems of Single—period thegry and those of Cf. the reprint in (1982).

29. A 'Robertson lag' in income is yet another possible bridge from 'week' to 'week'. In my (1968), I tried to get to the §§D§Cii Ibegry by this route: I had a first period in which sales declined because sellers had inelastic price expectations and thus did not cut prices fast enough; in the next week, demand was then 'income-constrained' with consequent Keynesian - effects, etc. I thought at- the time that I had, in effect, gotten over from Value and Capital to the General Theory in fairly good order and it puzzled me why Hicks had not tried this route. But Hicks had defined his Temporary Equilibrium in such a way as to preclude unintended shortfalls in sales. See his comments below on the "indefensible trick." 30. Economic Perspectives, p. vi.i. The sentence in quotes is from Q§Eit§! and Grgwth, Chapter VI, where the matter is also discussed. Compare also Clower [19753 and Clower and Lerjonhufvud C19753. "

31."Methods of Dynamic Analysis," section iv.

32. Ecgngmic Perspectives, p. xi. Cf. also Capital and Grgwth, Chapter VII. Mgneyj. iQterestj. and Wages, pp. 226-235, 296-99, 320-24.

33. Cf., "IS-LM: An Explanation," in Fitoussi, ed., C19833 and also included in Hicks C19823.

34. Cf. e.g., The Crisis in Keynesian Ecgngmics, p. 6. Also, "Recollections and Documents" in Economic Perspectives. This paper also records Keynes' detailed and favorable reaction to the IS-LM representation of his theory.

35.. Cf. , Critical Essays in Mgnetary. Ihegry, p. vi i: "But as a diagnosis of the 'revolution', CIS-LM3 is very unsatisfactory. It is not a bad representation of Keynes; but it does not get his predecessors (the 'Classics' as he called them) at all right."

36. Cf., e.g., "Time in Economics," in Hicks E19823, pp. 289-90: "All tVie same, I must say that that diagram is now much less popular with me than I think it still is with many other people. It reduces the General Thegry to equilibrium economics; it is not really in time. That, of course, is why it has done so well."

37. Approximately, from A Cgntriputign to the) Theory gf the Trade £y£ie (1950) to Capital and Grgwth (1965).' Or, perhaps, for the duration of his tenure as Drummond Professor (1952—65). For his preoccupations during this period, cf., "The Formation of an Economist," p. 202.

38. 0. Lange, C19423 and C19443. _

39. The book, he said, was written not "to elucidate the ',' but to deny that it is a revolution at all." Cf. Hicks C19573. This judgment was not fair to Patinkin as Hicks has acknowledged. Ci. his C1979c, n. 53.

40. Patinkin understood, of course, that this model would produce unemployment only if one imposed the restriction of rigid (and too high) wages. He also was quite clear on the fact that Keynes had assumed neither rigid wages nor a liquidity trap. (Patinkin had in any case demonstrated already in his C19483 that a liquidity trap would not by itself lead to unemployment in this type of model). Consequently, he chose to deal with Keynesian unemployment informally, discussing the unemployment dynamics of the system "off the curves" of his formal model. Cf. Patinkin C1956, Chapter 133.

41. Cf., "The 'Classics' Again," as reprinted in Hicks C19673, p. 143. My reasons for judging this to be the right point of departure are spelled out at great length "in "The Wi'cksell Connection" in Leijonhufvud C19813.

42. Cf. Hicks (1957), Patinkin (1959).

43. Alan Coddington discusses this Hicksian analysis in somewhat more' detail in his C1983, pp. 68-733.

44. Cf., Capital and Grgwth, p. 65. The particular difficulty ("...now lulled to sleep by long familiarity") mentioned in this context was that "CKeynes' theory3 works with a period which is taken to be one of equilibrium ... and which is nonetheless identified with the Marshal 1ian 'short period', in which capital equipment ... remains unchanged. The second seems to require that the period should not be too long, but the first that it should not be too short;... It is not easy to see that there can be any length of time that will adequately satisfy both of these requirements." (pp. 64-65). One notes that this observation would seem to threaten the legitimacy of Hicks' accordion playing with the period in his "The 'Classics' Again."

45. :Cri§is in Keynesian Ecgngmics, pp. 38-9.

46. "Time in Economics," in C19823, pp. 288-9.

47. ibid.., p. 290.

48. "IS-LM: An Explanation," in Fitoussi (1983), p. 57.

49. Causality in Ecgngmics, p. 85.

50. Cf.., "IS-LM: An Explanation", pp. 60-2. The brief summary in the text fails, I am afraid, to do justice to the sublety of Hicks' argument. The reader who would appraise it should consult also his Causality in Ecgngmics, Chapters VI and VII.

51. Lei jonhuf vud (1983b, p. 86). But the IS-LM interpretati-on of Keynes still has backers who feel that the algebra cannot but lead us right. (who has, of course, advocated the sticky wages view as preserving the essentials of Keynes' theory) sees preoccupation with the model's conceptual foundations as revealing some sort of anti—mathematical obscurantism. See his Keynes centennial article in The Ecgngmist, June 25, 1983.

52. The equivalence of the liquidity and loanable funds approaches to interest determination was argued by Hicks the Younger in his 1936 review of Keynes and in Value and Capital, Chapter XII. There the argument was made in a Temporary Equilibrium context but 'it has been carried over to IS-LM by others. The argument is, I think, misleading — except possibly in the context of rational expectations models; if the general equilibrium consequences of some parameter change are 'rationally anticipated' , all markets would 'open' with the new equilibrium prices already 'posted'. For silch a conceptual experiment, it indeed does not make sense to ask which excess demand was responsible for the change in which price. One must (to make sense) consider instances where, once price-setters have^posted prices based on their best forecasts, actual trading produces excess demands and supplies thus revealing the 'errors' in the forecasts. The issue of the versus loanable funds squabble is how this error-activated feedback control of price works in the case of the interest rate — specifically, whether the interest rate is 'ggyerned.1 by the excess demand for money or by the excess supply for securities. To discriminate between the two hypotheses, one must then consider states of the economy which do ngt have ED for money and ES of securities (or vice versa) at the same time. In a Keynes model, a "decline in MEC" produces an example, namely, a state with an ES of commodities and a corresponding ED for securities, while — at this stage of the seguence — the ED for money is still zero. If the loanable funds hypothesis is true, it is possible that the intertemporal price mechanism will take care of the intertemporal coordination problem (without a ); if the liquidity preference hypothesis is true, it is inconceivable.

53. For a more careful and detailed discussion, cf. Leijonhufvud, C1983b, pp.69-70, 76-80 3.

54. This sounds suspiciously like an IQ ranking for Lucasian, Friedmanian and Keynesian economists. This Keynesian didn't mean, it that way!

55. Cf., Leijonhufvud C19S3b, p. 873.

56. Cf., Critical Essays, p. 158.

57. I have made a previous attempt at getting Hicks' "Attempt at Perspective" into perspective — and pretty much failed. Cf., Leijonhufvud C19S1, Chapter 83. My review shows how influenced I then was by Friedman and Schwartz, Brunner and Meltzer and particularly by their work on United States monetary history since 1929. (In 1968, American monetarists had hardly began thinking about small, open, fixed exchange rate yet). This made me critical, for instance, of Hicks' insistence on the "inherent instability of credit." The piece also shows my great fascination for Hicks' daring attempt to put 200 years of tangled controversies in order; for various reasons, the way I saw it, several important writers just would not fit neatly into Hicks' scheme — but I failed completely to suggest a scheme that would do better.

58. The main line of Hicks' work in monetary theory - runs as follows: "A Suggestion for Simplifying the Theory of Money" (1935); Chapter XXIII, "Keynes After Growth Theory" in Capital Growth, (1965); the three chapters on "The Two Triads" in Critical Essays, (1967); the chapter on "Money, Interest and Liquidity" in The Crisis (1974); the 60-odd pages long "Monetary Experience and the Theory of Money" which is th.e backbone of the ioongmic Perspectives collection ~<1977); and "the Foundations of Monetary Theory" in MgneyA Interest and Wages (1982).

59. Quoted from reprint in Critical Essays, pp. 75-76.

60. Capital and Grgwth, pp. 284-5.

61. I have used this rather informal definition repeatedly. Cf.,e.g., Leijonhufvud C1983a3.

62. Cf., Leijonhufvud C19823 and C1983a3 for rather more careful explanations.

63. Cf."Monetary Experience and the Theory of Money," pp. 61-73, and "Foundations of Monetary Theory," pp. 237, and 264ff.

64. On which Hicks can rightfully say: "... I may allow myself to point out that it was already observed in 'Mr. Keynes and the Classics' that we do not need to suppose that the curve is drawn up on the assumption of a given stock of money. It is sufficient to suppose that there is (as I said) 'a given monetary system — that up to a point, but only up to a point, monetary authorities will prefer to create new money rather than allow interest rates to rise. Such a generalized (LM) curve will then slopfi. upwards only gradually — the of the curve depending on the elasticity of the monetary system...'." Cf., • Mgneyx Interest and Wl91§? P- 328.

65. Cf. Leijonhufvud (1983a).

66. Leijonhufvud (1979), p. 526.

67. Ecgngmig Perspectives, p. vii.

68. ibid.

69. "Monetary Theory and History," in Hicks (1967), p. 156^-But this too is an old Hicksian theme. One finds it in his 1943 review of Charles Rist's bistgry of Money and Credit Thegry. Cf. Hicks C19823, pp. 132 ff. REFERENCES

Coddington, Alan, (1983), Keynesian Ecgngmics:. The Search for First Principles, London: Allen ?<*Unwin.

Clower, Robert W. (1975), "Reflections on the Keynesian Perplex," 2e.itschrift fur Natignalgkgngmie.

Clower, Robert W. and , "The Coordination of Economic Activities: A Keynesian Perspective," American Ecgngmic Review, May.

Fitoussi, Jean-Paul, ed., (1983), Mgdern Macrgecgngmig Theory, Oxford: Blackwell.

Grandmont, Jean Michel, (1975), "Temporary general Equilibrium Theory," Ecgngmetrica.

Hart, Albert G. (1942), "Risk, Uncertainty, and the Unprofitabi1ity of Compounding Probabilities," in Oscar Lange, et al. , Studies in and , Chicago. Reprinted in W. Fellner and B.F. Haley, eds.,C19513 Readings in the Thegry of income Distrigutign, Philadelphia: Blakeston.

Hayek, Friedrich A. von (1928), " "Das intertemporale Gleichgewichtssystem," Weitwirtschaftiiches Archiy.

Hayek, Fjriedrich A. von, ed. , (1933), Beitrage zur Geidthegrie.

Heiner, Ron, (1983), "The Origin of Predictable Behavior," Ecgngmic Review, September.

Hicks, J.R. and R.G.D. Allen,(1934), "A Reconsideration of the Theory of Value: Parts I and II," Ecgngmica N^S^

Hicks, Sir John, (1939), Value and Capital, Oxford: Oxford University Press.

Hicks, Sir John, (1956), "Methods of Dynamic Analysis," in 25 Ecgngmic Essays in Hgngur g£ Eri]< Lindahl, ed. by the eds.— of Ekgngmisk lidskrift, Stockholm.

Hicks, Sir John, (1957), "A Rehabilitation of 'Classical' Economics?" Ecgngmic Jgurnal, June.

Hicks, Sir John, (1965), Capital and Grgwth, Oxford: Oxford University Press.

Hicks, Sir John, (1967), Critical Essays in Mgnetary Theory, Oxford: Oxford University Press.

Hicks, Sir John, (1974), The Crisis in Keynesian Oxford: Blackwell. Hicks, Sir John, (1974), The Crisis in Keynesian Ecgngmics, Oxford: Blackwell. Hicks, Sir John, (1977), Economic "Perspectives! ""Further Essays on Money and Grgwth, Oxford: Oxford University Press. v. Hicks, Sir John, (1979a), Causality in Economics, Oxford: Blackwell. Hicks, Sir John, (1979b), "The Formation of an Economist," Banca N§.z.ignale del L^yorg Quarterly Beyiew, September.

Hicks, Sir John, (1979c), "On Coddington's Interpretation: A Reply," Journal of Ecgngmic Literature, September.

Hicks, Sir John, (1981), Wealth and Welfare! CgHected Essays on Ecgngmic Thegryj. Vol. I, Oxford: Blackwell.

Hicks, Sir John, (1982), Mgneyj. Interest and Wages! Collected on Economic Thegry, Vol. II, Oxford: Blackwell'.

Hicks, Sir John, (1983), Classics and Moderns! Collected Essays on Ecgngmic Thegry, Vol. Ill, Oxford: Blackwell.

Lange, Oscar, C19423, "Say's Law: A Restatement and Criticism," in 0. Lange et aiij. Studies in Mathematical Ecgngmics and Ecgngmetrigs, Chicago: University of Chicago Press.

Lange, Oscar, C19443, Pri.ce Flexibility and Full Employment, Chicago: Cowles Commission.

Leijonhufvud, Axel, (1968), On Keynesian Ecgngmics and the Ecgngmics gf Keynes, New York: Oxford.

Leijonhufvud, Axel, (1979) "Review of Ecgngmic Perspectives by Sir John Hicks," Jgurnai gf Ecgngmig Literature, June, pp. 525-

4. / •

Leijonhufvud, Axel, (1981), lnf_grmatign and Coordination! Essays ID Macrgecgngmic Thegry, New York: Oxford University Press. —

Leijonhufvud, Axel, (1982), "Rational Expectations and Monetary Institutions," paper presented at the International Economic Association Conference on Mgnetary Thegry and Monetary Institutions, Florence, Italy, September. Leijonhufvud, Axel, (1983a), "What Would Keynes Have Thought of Rational Expectations?" in G.D.N. Worswick and J.S. Trevithick, eds., Keynes and the Modern World, Cambridge: Cambridge University Press.

Leijonhufvud, Axel, (1983b), "What Was the Matter with IS-LM?" in Jean-Paul Fitoussi, ed., "' Modern Macrgecgngmic Theory! An w, Oxford: Blackwell. Marshall, Alfred, (1928), Principles gf Ecgngmics, Eighth edn., London: MacMillan.

Myrdal, Gunnar, (1933), "Geldtheoretisches -Gleichgewicht," in Hayek, ed.,(1933). *

Patinkin, Don, C19483, "Price Flexibility and Full Employment," Ecgngmic Review, September.

Patinkin, Don, C1956 3, MgneyA Interestj. and Prices, Evans ton, 111.: Row, Peterson & Co.

Patinkin, Don, C19593, "Keynesian Economics Rehabilitated: A Rejoinder to Professor Hicks," Economic Jgurnai, September.

Robbins, Lionel, (1932), An Essay on the Nature and Significance of Ecgngmic Science, London: MacMillan.

Shackle, G.L.S., (1967), The Years gf High Theory! Invention and Ir§^iii5D iD Economic Jhgughtj. 1926-1939, Cambridge: Cambridge University Press.

Shackle, G.L.S., (1972), Epistegjics and Ecgngmics! A Critigue gf Dgctrines, Cambridge: Cambridge University Press.